myaiforone 1.1.47 → 1.1.49

package/.claude/commands/setup.md

@@ -243,6 +243,33 @@ EOF
 
 cat > "$APP_PATH/Contents/MacOS/AIforOne" << 'EOF'
 #!/bin/bash
+export PATH="/usr/local/bin:/opt/homebrew/bin:/usr/bin:/bin:$PATH"
+
+# Check if already running
+if curl -s --max-time 2 http://localhost:4888/health 2>/dev/null | grep -q "ok"; then
+    open "http://localhost:4888"
+    exit 0
+fi
+
+# Not running — try launchctl first
+PLIST="$HOME/Library/LaunchAgents/com.agenticledger.channelToAgentToClaude.plist"
+if [ -f "$PLIST" ]; then
+    launchctl load "$PLIST" 2>/dev/null
+else
+    # Fallback: start via npx (detached)
+    nohup npx myaiforone@latest > /dev/null 2>&1 &
+fi
+
+# Wait for health (up to 30s)
+for i in $(seq 1 30); do
+    if curl -s --max-time 2 http://localhost:4888/health 2>/dev/null | grep -q "ok"; then
+        open "http://localhost:4888"
+        exit 0
+    fi
+    sleep 1
+done
+
+# Timeout — open anyway, user will see loading
 open "http://localhost:4888"
 EOF
 

package/agents/platform/gym/CLAUDE.md

@@ -46,15 +46,15 @@ No clarifying questions. Execute immediately using the specified tools.
 ### Guide Generation (all modes)
 After any substantive session: `create_gym_guide` with title, description, steps, dimensions, difficulty. Ask user to review. Publish as skill via `create_skill` when appropriate.
 
-## The 5 Dimensions (1–5 scale, 0 = unassessed)
+## The 5 Dimensions — "Using AI for:" (1–5 scale, 0 = unassessed)
 
-| Dimension | Measures |
-|-----------|----------|
-| **Application** | Using agents for real work, right agent for job, iterating on results |
-| **Communication** | Prompt quality, context loading, course correction, prompt evolution |
-| **Knowledge** | Understands agents/tools/MCPs/memory conceptually, can troubleshoot |
-| **Orchestration** | Multi-agent workflows, cron/goals, projects, delegation chains |
-| **Craft** | Creates/tunes agents: system prompts, tool curation, MCPs, workspaces |
+| Dimension | Key | Measures |
+|-----------|-----|----------|
+| **Gaining Knowledge** | knowledge | Using AI to research, find information, learn concepts, evaluate answers |
+| **Communication** | communication | Prompting AI effectively, using AI to write for others, multi-channel content |
+| **Analysis, Reviews and Validation** | analysis | Using AI to check work, compare options, validate data, review documents |
+| **Automating Manual Tasks** | automation | Delegating repetitive tasks, multi-step workflows, goals/cron, orchestration |
+| **Building Applications, Tools and Dashboards** | building | Creating tools, apps, dashboards with AI; agent configs, MCP integrations |
 
 Assess from observed activity, not self-report. Call `snapshot_dimensions` after any score update.
 
@@ -86,15 +86,15 @@ Same full MCP access as @hub — agents, tasks, projects, automations, skills, M
 
 | Gap | Recommend |
 |-----|-----------|
-| Low Application (<2) | On-the-job training with real tasks |
-| Low Communication (<2) | Prompt Engineering program |
 | Low Knowledge (<2) | Getting Started program |
-| Low Orchestration (<2) | Automations Mastery program |
-| Low Craft (<2) | Agent Building program |
+| Low Communication (<2) | Prompt Engineering program |
+| Low Analysis (<2) | On-the-job training: ask AI to review real documents |
+| Low Automation (<2) | Automations Mastery program |
+| Low Building (<2) | Agent Building program |
 | All low | Getting Started first, then reassess |
 | All 3+ | Advanced programs or on-the-job challenges |
 
-Priority order when multiple gaps: Knowledge → Application → Communication → Craft → Orchestration
+Priority order when multiple gaps: Knowledge → Communication → Analysis → Automation → Building
 
 ## Verification Methods
 
@@ -138,11 +138,11 @@ Heuristic digest (6am) handles scoring by rules and template cards. Your job: ac
 7. `get_gym_progress` — program completion
 
 ### Score each dimension:
-- **Application**: real work vs. test messages, right agent used, results iterated, conversations concluded
-- **Communication**: prompt specificity, context loaded, correction quality, prompt evolution over time
 - **Knowledge**: correct concept use, feature awareness, troubleshooting ability, program completion depth
-- **Orchestration**: active automations (lastRun exists), cross-agent workflows, project usage, cron sophistication
-- **Craft**: system prompt quality, tool curation intentionality, MCP fit to purpose, workspace specificity, design iteration
+- **Communication**: prompt specificity, context loaded, correction quality, prompt evolution over time
+- **Analysis**: using AI to review/validate work, structured analysis, data checking, QA patterns
+- **Automation**: active automations (lastRun exists), cross-agent workflows, project usage, cron sophistication
+- **Building**: system prompt quality, tool curation intentionality, MCP fit to purpose, workspace specificity, design iteration
 
 ### Synthesize:
 - Note where your scores differ from heuristic and why

package/agents/platform/gym/ai-gym-curriculum-framework.md

@@ -0,0 +1,101 @@
+# AI Gym Curriculum Framework
+
+## Philosophy
+
+The goal is not to learn AI — the goal is to **use AI**. Every dimension answers: "What did you USE AI to DO?"
+
+Each module progresses through 4 levels (university model):
+- **101** = You + AI, one-shot, manual
+- **201** = You + AI, multi-step, structured
+- **301** = AI runs it with your design, automated
+- **401** = AI systems that operate independently, production-grade
+
+---
+
+## Module 1: Using AI to Gain Knowledge & Find Information
+
+| Level | Core Idea | Focus |
+|---|---|---|
+| **101** | Ask AI questions | Chat with AI to get answers, replace basic Googling, learn to evaluate AI responses |
+| **201** | Research with AI | Multi-step research, comparing sources, follow-up questioning, summarizing long documents |
+| **301** | Structured knowledge systems | RAG, knowledge bases, feeding AI your own data, building searchable archives |
+| **401** | Autonomous knowledge agents | Agents that monitor, curate, and surface information proactively (news digests, competitive intel, trend alerts) |
+
+---
+
+## Module 2: Using AI to Communicate
+
+| Level | Core Idea | Focus |
+|---|---|---|
+| **101** | Talk to AI effectively | Prompting fundamentals — clear instructions, context-giving, iterating on outputs |
+| **201** | Use AI to write for others | Draft emails, proposals, reports, social posts — AI as your writing partner |
+| **301** | Multi-channel AI communication | AI across Slack, email, LinkedIn, client-facing docs — tone/audience adaptation, templates |
+| **401** | AI as communication strategist | Content calendars, audience segmentation, brand voice systems, automated outreach pipelines |
+
+**Note:** This dimension intentionally covers BOTH communicating WITH AI (101) and using AI to communicate WITH OTHERS (201+).
+
+---
+
+## Module 3: Using AI to Analyze, Review & Validate
+
+| Level | Core Idea | Focus |
+|---|---|---|
+| **101** | Ask AI to check your work | Paste in a doc/spreadsheet/email, ask "what's wrong?" — basic review and proofreading |
+| **201** | Structured analysis | Compare options, pros/cons, financial reviews, data validation, gap analysis |
+| **301** | AI-powered QA workflows | Checklists, compliance checks, audit prep, systematic review processes with AI |
+| **401** | Autonomous validation pipelines | Agents that continuously monitor, flag anomalies, and validate data integrity across systems |
+
+---
+
+## Module 4: Using AI to Improve Manual Tasks (Process Execution)
+
+| Level | Core Idea | Focus |
+|---|---|---|
+| **101** | Delegate a task to AI | Hand AI a single repetitive task — reformatting, data entry, cleanup, sorting |
+| **201** | Multi-step process delegation | Chain steps together — AI handles a full workflow (e.g., reconcile → categorize → report) |
+| **301** | Orchestration | Goals, scheduled tasks, cron jobs — AI runs processes on autopilot without you asking |
+| **401** | Multi-agent process systems | Multiple agents collaborating on complex processes, delegation chains, exception handling |
+
+**The test:** If you delete it after it runs, it was Module 4 (process). If it lives on and gets reused, it's Module 5 (asset).
+
+---
+
+## Module 5: Using AI to Build Applications, Tools & Dashboards (Asset Building)
+
+| Level | Core Idea | Focus |
+|---|---|---|
+| **101** | Build a simple tool with AI | Use AI to create a calculator, template, or simple utility — your first "vibe code" |
+| **201** | Build functional apps | Full applications with UI, data, and logic — AI as your developer |
+| **301** | Integrate and connect | MCP configs, API integrations, tools that pull live data from real systems |
+| **401** | Production systems | Deployed apps, multi-user tools, monitoring, maintenance — assets that run your business |
+
+---
+
+## Curriculum Grid
+
+5 modules x 4 levels = **20 course slots**. Each slot contains 5 programs = **100 programs at full buildout**.
+
+---
+
+## Mapping from Old Dimensions
+
+| Old Dimension | New Home | Notes |
+|---|---|---|
+| Knowledge | Module 1 | Direct fit |
+| Communication | Module 2 | Shifted: was "prompting skill", now covers both prompting (101) and communicating with others (201+) |
+| Application | Dissolved | Becomes an engagement metric across all 5, not its own dimension |
+| Craft | Module 5 | Building things |
+| Orchestration | Module 4 at 301+ | Advanced process execution |
+| Analyze/Review | Module 3 | **Net new** — no old dimension covered this |
+
+---
+
+## Scoring Signals (for dimension-scorer)
+
+| Dimension Key | Label | What to Measure |
+|---|---|---|
+| `knowledge` | Gaining Knowledge | Program completions + gym engagement + topic breadth |
+| `communication` | Communication | Messages involving content creation tools (Write, Edit) + communication MCPs (Slack, email, LinkedIn) |
+| `analysis` | Analysis, Reviews & Validation | Read/Grep/Glob tool usage + review-oriented conversations + data MCPs (Plaid, QBO) |
+| `automation` | Automating Manual Tasks | Cron/goals setup + automation patterns + repeated task delegation |
+| `building` | Building Apps, Tools & Dashboards | Agent creation + MCP configs + system prompts + workspace setups + Bash usage |

package/agents/platform/gym/programs/32-claude-shortcut-hacks-for-faster-prompts/program.json

@@ -0,0 +1,57 @@
+{
+  "id": "32-claude-shortcut-hacks-for-faster-prompts",
+  "slug": "32-claude-shortcut-hacks-for-faster-prompts",
+  "title": "32 Claude Shortcut Hacks for Faster Prompts",
+  "description": "A practical training program to master Claude prompt shortcuts.",
+  "coverImage": null,
+  "sourceUrl": null,
+  "tier": "free",
+  "personas": [],
+  "globalInfo": null,
+  "tags": [],
+  "isActive": true,
+  "isPublic": true,
+  "source": "platform",
+  "sourceId": "dbaf1d3f-608d-4b2b-ac26-0f866eeab39d",
+  "projectSlug": "public",
+  "seriesSlug": null,
+  "orderInSeries": null,
+  "difficulty": "beginner",
+  "dimensions": [],
+  "estimatedTime": null,
+  "prerequisites": [],
+  "trainers": [
+    "alex",
+    "jordan",
+    "morgan",
+    "riley",
+    "sam"
+  ],
+  "createdAt": "2026-04-10T03:44:01.502Z",
+  "updatedAt": "2026-04-19T01:46:03.406Z",
+  "modules": [
+    {
+      "id": "module-0",
+      "title": "Simplify and Summarize",
+      "description": "Learn to instantly control how complex or concise a response is.",
+      "order": 0,
+      "agentInstructions": null,
+      "isVisible": true,
+      "steps": [
+        {
+          "id": "step-0-0",
+          "title": "/ELI5 - Explain Like I'm 5",
+          "order": 0,
+          "type": "self-report",
+          "content": "Add /ELI5 at the start of any prompt to get a simple, jargon-free explanation.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        }
+      ]
+    }
+  ]
+}

package/agents/platform/gym/programs/7-claude-cowork-skills-every-accountant-should-install/program.json

@@ -0,0 +1,154 @@
+{
+  "id": "7-claude-cowork-skills-every-accountant-should-install",
+  "slug": "7-claude-cowork-skills-every-accountant-should-install",
+  "title": "7 Claude Cowork Skills Every Accountant Should Install",
+  "description": "Install 7 ready-made Claude skills that handle common accounting workflows — from tax workpapers to depreciation schedules.",
+  "coverImage": null,
+  "sourceUrl": null,
+  "tier": "free",
+  "personas": [],
+  "globalInfo": null,
+  "tags": [
+    "AI",
+    "Accounting",
+    "Claude",
+    "Skills",
+    "Automation"
+  ],
+  "isActive": true,
+  "isPublic": true,
+  "source": "platform",
+  "sourceId": "25f48ef6-ccbe-47fb-8bb7-be03ca22b568",
+  "projectSlug": "unofficial-guides-from-mohamed-muse-said-s-li-posts",
+  "seriesSlug": null,
+  "orderInSeries": null,
+  "difficulty": "beginner",
+  "dimensions": [],
+  "estimatedTime": null,
+  "prerequisites": [],
+  "trainers": [
+    "alex",
+    "jordan",
+    "morgan",
+    "riley",
+    "sam"
+  ],
+  "createdAt": "2026-04-19T01:53:40.623Z",
+  "updatedAt": "2026-04-19T01:53:40.623Z",
+  "modules": [
+    {
+      "id": "66a6ae24-c8ac-463d-b4fa-90527fafe5ca",
+      "title": "Steps",
+      "description": "",
+      "order": 0,
+      "agentInstructions": null,
+      "isVisible": true,
+      "steps": [
+        {
+          "id": "27269d7d-64d9-4d37-9744-93e08119f06f",
+          "title": "Understand what Claude Skills are and why they matter",
+          "order": 1,
+          "type": "self-report",
+          "content": "Skills are reusable markdown files that give Claude pre-built instructions for specific tasks. Instead of re-prompting every time, you load a skill and it executes immediately. Less context to send, fewer tokens used, faster and more consistent results.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "5d48cc68-506c-482a-8581-2c9d996d6995",
+          "title": "Install the Month-End Close Agent skill",
+          "order": 2,
+          "type": "self-report",
+          "content": "This skill handles the full close pipeline: pull data, reconcile, draft entries, produce statements, write commentary, package workpapers, and finalize the close memo. Run the whole pipeline or just the piece you need.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "f7121f9c-fdba-41db-87cf-ff498461f25c",
+          "title": "Install the Tax Workpaper Populator skill",
+          "order": 3,
+          "type": "self-report",
+          "content": "This skill takes source tax documents and populates your workpaper template automatically. It reads through PDFs, extracts the relevant data points, and drops them into the right fields.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "294e09a6-199a-4783-ab70-46c801febb98",
+          "title": "Install the Budget vs. Actuals Reporter skill",
+          "order": 4,
+          "type": "self-report",
+          "content": "This skill pulls your budget and actuals, calculates variances, and writes commentary explaining what drove each significant difference. Output is ready for management review.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "22ac24a6-d4d9-42f4-955e-d1c7f3aed898",
+          "title": "Install the Missing Documents Tracker skill",
+          "order": 5,
+          "type": "self-report",
+          "content": "This skill cross-references your client checklist against what's been received. It identifies what's still missing, who needs to provide it, and drafts follow-up messages.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "f42465fd-dce1-4f42-8a45-c820f87051de",
+          "title": "Install the Bank Reconciliation Builder skill",
+          "order": 6,
+          "type": "self-report",
+          "content": "A standalone reconciliation skill that matches bank statement transactions to GL entries, flags unmatched items, and produces a formatted reconciliation workpaper.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "8e51b143-446d-4034-81e0-b11fa79001d7",
+          "title": "Install the Invoice & Receipt Extractor skill",
+          "order": 7,
+          "type": "self-report",
+          "content": "This skill reads invoices and receipts (PDF or image), extracts vendor, amount, date, and category, and outputs a structured table ready for entry or review.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "dac50542-85f5-4240-b149-d94125b4bded",
+          "title": "Install the Depreciation Scheduler skill",
+          "order": 8,
+          "type": "self-report",
+          "content": "This skill takes your fixed asset register, applies the correct depreciation method and useful life for each asset, and produces the monthly depreciation schedule with journal entries.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        }
+      ]
+    }
+  ]
+}

package/agents/platform/gym/programs/accounting-firm-20-to-35-margin/program.json

@@ -0,0 +1,57 @@
+{
+  "id": "accounting-firm-20-to-35-margin",
+  "slug": "accounting-firm-20-to-35-margin",
+  "title": "Take Your Accounting Firm from 20% to 35% Margin",
+  "description": "Follow the same framework Jason Staats used to increase his $4M firm's margin from 20% to 35% in 18 months — by fixing your client list, pricing, and focus. No tech tricks — just hard business decisions.",
+  "coverImage": null,
+  "sourceUrl": null,
+  "tier": "free",
+  "personas": [],
+  "globalInfo": "Source: Jason Staats, CPA — LinkedIn post (2026-04-18 scrape). Core thesis: your firm's margin problem is a client list problem, not a tech problem. Fix clients first, then systems, then tech. Move low-margin clients along, use multi-tier pricing to reveal who wants to pay more, and find better clients through content and niche positioning.",
+  "tags": [
+    "Accounting",
+    "Firm Management",
+    "Pricing",
+    "Profitability"
+  ],
+  "isActive": true,
+  "isPublic": true,
+  "source": "platform",
+  "sourceId": "006d8429-b588-4777-a375-70a62f220a54",
+  "projectSlug": "unofficial-guides-from-josh-staats-li-posts",
+  "seriesSlug": null,
+  "orderInSeries": null,
+  "difficulty": "beginner",
+  "dimensions": [],
+  "estimatedTime": null,
+  "prerequisites": [],
+  "trainers": [
+    "alex",
+    "jordan",
+    "morgan",
+    "riley",
+    "sam"
+  ],
+  "createdAt": "2026-04-19T01:52:47.261Z",
+  "updatedAt": "2026-04-19T01:52:47.261Z",
+  "modules": [
+    {
+      "id": "module-0",
+      "title": "Diagnose the Problem",
+      "description": "",
+      "order": 1,
+      "agentInstructions": null,
+      "isVisible": true,
+      "steps": []
+    },
+    {
+      "id": "module-1",
+      "title": "Fix and Grow",
+      "description": "",
+      "order": 2,
+      "agentInstructions": null,
+      "isVisible": true,
+      "steps": []
+    }
+  ]
+}

package/agents/platform/gym/programs/ai-proof-your-accounting-firm-s-tech-stack/program.json

@@ -0,0 +1,128 @@
+{
+  "id": "ai-proof-your-accounting-firm-s-tech-stack",
+  "slug": "ai-proof-your-accounting-firm-s-tech-stack",
+  "title": "AI-Proof Your Accounting Firm's Tech Stack",
+  "description": "Understand what makes software future-proof in an AI-agent world.",
+  "coverImage": null,
+  "sourceUrl": null,
+  "tier": "free",
+  "personas": [],
+  "globalInfo": null,
+  "tags": [
+    "AI",
+    "Accounting",
+    "Tech Stack",
+    "MCP",
+    "Future-Proofing"
+  ],
+  "isActive": true,
+  "isPublic": true,
+  "source": "platform",
+  "sourceId": "e4b7d8f8-bd49-4421-bcf8-18bf96a3db5a",
+  "projectSlug": "public",
+  "seriesSlug": null,
+  "orderInSeries": null,
+  "difficulty": "beginner",
+  "dimensions": [],
+  "estimatedTime": null,
+  "prerequisites": [],
+  "trainers": [
+    "alex",
+    "jordan",
+    "morgan",
+    "riley",
+    "sam"
+  ],
+  "createdAt": "2026-04-18T19:41:40.762Z",
+  "updatedAt": "2026-04-19T01:46:03.407Z",
+  "modules": [
+    {
+      "id": "module-0",
+      "title": "Steps",
+      "description": "Evaluate and future-proof your tech stack.",
+      "order": 0,
+      "agentInstructions": null,
+      "isVisible": true,
+      "steps": [
+        {
+          "id": "step-0-0",
+          "title": "Understand the two jobs that remain for software",
+          "order": 0,
+          "type": "self-report",
+          "content": "Build AI agents or build software for AI agents.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "step-0-1",
+          "title": "Look for MCP server support as the gold standard",
+          "order": 1,
+          "type": "self-report",
+          "content": "MCP lets AI agents interact with your software directly.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "step-0-2",
+          "title": "At minimum, demand a good API",
+          "order": 2,
+          "type": "self-report",
+          "content": "Tools should have a robust API for programmatic access.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "step-0-3",
+          "title": "Evaluate your current tools",
+          "order": 3,
+          "type": "self-report",
+          "content": "Ask: Can an AI agent interact with this tool?",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "step-0-4",
+          "title": "Prioritize connectivity over features",
+          "order": 4,
+          "type": "self-report",
+          "content": "Weight connectivity as heavily as features when evaluating new tools.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        },
+        {
+          "id": "step-0-5",
+          "title": "Start testing agent workflows with one tool today",
+          "order": 5,
+          "type": "self-report",
+          "content": "Pick one tool, connect it to an AI agent, give it a simple task.",
+          "isCritical": false,
+          "personaVariations": {},
+          "trainerVariations": {},
+          "attachments": [],
+          "isVisible": true,
+          "verification": "self-report"
+        }
+      ]
+    }
+  ]
+}